Exact Computation of Influence Spread by Binary Decision Diagrams

Total Page:16

File Type:pdf, Size:1020Kb

Exact Computation of Influence Spread by Binary Decision Diagrams Exact Computation of Influence Spread by Binary Decision Diagrams Takanori Maehara1;2) Hirofumi Suzuki3) Masakazu Ishihata3) 1) Shizuoka University 2) RIKEN Center for Advanced Intelligence Project 3) Hokkaido University [email protected] [email protected] [email protected] ABSTRACT 1. INTRODUCTION Evaluating influence spread in social networks is a funda- mental procedure to estimate the word-of-mouth effect in 1.1 Background and Motivation viral marketing. There are enormous studies about this Viral marketing is a strategy to promote products by giv- topic; however, under the standard stochastic cascade mod- ing free (or discounted) items to a selected group of highly in- els, the exact computation of influence spread is known to be fluential individuals (seeds), in the hope that through word- #P-hard. Thus, the existing studies have used Monte-Carlo of-mouth effects, a significant product adoption will occur [8, simulation-based approximations to avoid exact computa- 27]. To maximize the number of adoptions, Kempe, Klein- tion. berg, and Tardos [19] mathematically formalized the dy- We propose the first algorithm to compute influence spread namics of information propagation, and proposed the op- exactly under the independent cascade model. The algo- timization problem, referred to as the influence maximiza- rithm first constructs binary decision diagrams (BDDs) for tion problem. Several cascade models have been proposed, all possible realizations of influence spread, then computes and the most commonly used one is the independent cascade influence spread by dynamic programming on the constructed model, proposed by Goldberg, Libai, and Muller [10, 11]. BDDs. To construct the BDDs efficiently, we designed a In this model, the individuals are affected by information new frontier-based search-type procedure. The constructed that is stochastically and independently propagated along BDDs can also be used to solve other influence-spread re- edges in the network from the seed (Section 2.1). To date, lated problems, such as random sampling without rejection, significant efforts have been devoted to the development conditional influence spread evaluation, dynamic probability of efficient algorithms for the influence maximization prob- update, and gradient computation for probability optimiza- lem [1, 4{7, 25, 26, 31]. tion problems. Here we consider the computational complexity of the We conducted computational experiments to evaluate the influence maximization problem. Under the independent proposed algorithm. The algorithm successfully computed cascade model, the expected size of influence spread is a influence spread on real-world networks with a hundred edges non-negative submodular function [19]; thus, a (1 − 1=e) in a reasonable time, which is quite impossible by the naive approximate solution can be obtained by using a greedy al- algorithm. We also conducted an experiment to evaluate gorithm [24]. However, the evaluation of influence spread the accuracy of the Monte-Carlo simulation-based approxi- is #P-hard [4] because it contains the problem of count- mation by comparing exact influence spread obtained by the ing s-t connected subgraphs [32]. Thus all existing studies proposed algorithm. avoided the exact computation and employed the Monte- Carlo simulation-based approximation, which simulates the Keywords dynamics of information propagation sufficiently many times (e.g., Ω(1/2)) to obtain an accurate (e.g., 1 ± ) approxi- viral marketing; influence spread; enumeration algorithm; mation of influence spread [25] (Section 6). binary decision diagram In this study, we first tackle the problem of computing in- fluence spread exactly under the independent cascade model. As the problem is #P-hard, we are interested in an algo- rithm that runs on small real-world networks (i.e., having a few hundred edges) in a reasonable time. The motivations for this studies are as follows. • Influence spread over small networks is practically im- c 2017 International World Wide Web Conference Committee (IW3C2), portant. Because real social networks often consist of published under Creative Commons CC BY 4.0 License. many small communities, it is reasonable to consider WWW 2017, April 3-7, 2017, Perth, Australia. each community separately or consider only the inter- ACM 978-1-4503-4913-0/17/04. community network. http://dx.doi.org/10.1145/3038912.3052567 • When we wish to rank vertices according to their in- fluence spread, we need to compute the values accu- . rately. Monte-Carlo simulation cannot be used for this 947 purpose because it requires Ω(1/2) samples for 1 ± Let G = (V; E) be a directed graph with vertices V and approximation; thus < 10−5 is impossible. On the edges E. Each edge e 2 E has activation probability p(e). other hand, an exact method can be used because its Each vertex is either active or inactive. Note that inactive complexity does not depend on the desired accuracy. vertices may become active, but not vice versa. Here, an active vertex is considered \influenced." • Exact influence spread helps to analyze the quality of Suppose that information is propagated from S ⊆ V , Monte-Carlo simulation. Although many experiments which is called seeds. Initially, all vertices are inactive. using Monte-Carlo simulation have been conducted, Then, propagation over the network is performed as follows. none have been compared with the exact value because First, each seed u 2 S is activated. When u first becomes there is no algorithm that can compute this value. active, it is given a single chance to activate each currently • Establishing a practical algorithm for the fundamental inactive neighbor v with probability p((u; v)). This process #P-hard problem is interesting and important task in is repeated until no further activations are possible. The computer science. expected number of activated vertices after the end of the process is called influence spread, which is denoted as σ(S). 1.2 Contributions There is a useful interpretation of influence spread with In this study, we provide the following contributions. this model. We select each edge e 2 E with probability p(e). Then, we obtain edge set F . We then consider the induced • We propose an algorithm to compute influence spread subgraph G[F ] = (V; F ), which is a network consisting of exactly under the independent cascade model. Note only the selected edges. Here, let σ(S; F ) be the number that this is the first attempt to compute this value of vertices reachable from some u 2 S on G[F ]. Then, we exactly (Section 3). obtain the following: • The proposed algorithm enumerates all spread pat- X terns using binary decision diagrams (BDDs). Then, σ(S) = E[σ(S; F )] = σ(S; F )p(F ) (1) it computes influence spread by dynamic programming F ⊆E on the BDDs. Here, we have designed a new frontier- based search method, which constructs the BDD for where s-t connected subgraphs efficiently (Section 3.2). This Y Y 0 is the main technical contribution of this study. p(F ) = p(e) (1 − p(e )): (2) • We conducted computational experiments to evaluate e2F e02EnF the proposed algorithm (Section 5). We obtained the exact influence on real-world and synthetic networks We use this formula to compute the influence spread. with a hundred edges in reasonable times. We also compared the obtained exact influence with the one 2.2 Binary decision diagram obtained using the Monte-Carlo simulation. As discussed in Section 3, the exact evaluation of (1) in- In addition, using the constructed BDDs, we can also solve volves enumerating S-t connecting subgraphs, which is the the following influence-spread related problems (Section 4). graph having a path from S to t. To maintain exponentially many such subgraphs, we use the binary decision diagram • Random sampling from the set of realizations that suc- (BDD), which is a data structure to represent a Boolean cessfully propagates information helps to understand function compactly based on Shannon decomposition. Note the route of influence spread. We can perform this that a Boolean function can be used to represents set family without rejection by using the BDD. as the indicator function. A BDD is a directed acyclic graph D = (N ; A) with node • The conditional expectation of the influence spread set N and arc set A.1 It has two terminals 0 and 1. Each under the influenced (and non-influenced) conditions non-terminal node α 2 N is associated with variable e 2 E, on some vertices can be used to measure the effect of and has two arcs called 0-arc and 1-arc. The nodes pointed conducted viral promotion from a small observations. by 0-arc and 1-arc are referred to as 0-child and 1-child (de- This value is efficiently computed by the BDDs. noted by α0 and α1), respectively. A BDD represents a • When the activation probability changes, we can effi- Boolean function as follows: A path from the root node to ciently update the influence spread. the 1-terminal represents a (possibly partial) variable assign- ment for which the represented Boolean function is True. As • The derivatives of the influence spread with respect to the path descends to a 0-arc (1-arc) from a node, the node's the activation probabilities can be computed. This is variable is assigned to False (True). used to implement a gradient method for the influence A special type of BDD, i.e., reduced ordered binary deci- spread optimization problem. sion diagram (ROBDD) [3], is frequently used in practice. A BDD is ordered if different variables appear in the same 2. PRELIMINARIES order on all paths from the root. A BDD is reduced if the 2.1 Independent Cascade Model for Influence Spread 1To avoid confusion, we use the terms \vertex" and "edge" to refer to a vertex and edge in the original graph G, and The independent cascade model [10, 11] is the most com- \node" and \arc" to refer to a vertex and edge in the BDD monly used stochastic cascade model used for social network D.
Recommended publications
  • Studies on Enumeration of Acyclic Substructures in Graphs and Hypergraphs (グラフや超グラフに含まれる非巡回部分構造の列挙に関する研究)
    CORE Metadata, citation and similar papers at core.ac.uk Provided by Hokkaido University Collection of Scholarly and Academic Papers Studies on Enumeration of Acyclic Substructures in Graphs and Hypergraphs (グラフや超グラフに含まれる非巡回部分構造の列挙に関する研究) Kunihiro Wasa February 2016 Division of Computer Science Graduate School of Information Science and Technology Hokkaido University Abstract Recently, due to the improvement of the performance of computers, we can easily ob- tain a vast amount of data defined by graphs or hypergraphs. However, it is difficult to look over all the data by mortal powers. Moreover, it is impossible for us to extract useful knowledge or regularities hidden in the data. Thus, we address to overcome this situation by using computational techniques such as data mining and machine learning for the data. However, we still confront a matter that needs to be dealt with the expo- nentially many substructures in input data. That is, we have to consider the following problem: given an input graph or hypergraph and a constraint, output all substruc- tures belonging to the input and satisfying the constraint without duplicates. In this thesis, we focus on this kind of problems and address to develop efficient enumeration algorithms for them. Since 1950's, substructure enumeration problems are widely studied. In 1975, Read and Tarjan proposed enumeration algorithms that list spanning trees, paths, and cycles for evaluating the electrical networks and studying program flow. Moreover, due to the demands from application area, enumeration algorithms for cliques, subtrees, and subpaths are applied to data mining and machine learning. In addition, enumeration problems have been focused on due to not only the viewpoint of application but also of the theoretical interest.
    [Show full text]
  • Enumeration of Maximal Cliques from an Uncertain Graph
    1 Enumeration of Maximal Cliques from an Uncertain Graph Arko Provo Mukherjee #1, Pan Xu ∗2, Srikanta Tirthapura #3 # Department of Electrical and Computer Engineering, Iowa State University, Ames, IA, USA Coover Hall, Ames, IA, USA 1 [email protected] 3 [email protected] ∗ Department of Computer Science, University of Maryland A.V. Williams Building, College Park, MD, USA 2 [email protected] Abstract—We consider the enumeration of dense substructures (maximal cliques) from an uncertain graph. For parameter 0 < α < 1, we define the notion of an α-maximal clique in an uncertain graph. We present matching upper and lower bounds on the number of α-maximal cliques possible within a (uncertain) graph. We present an algorithm to enumerate α-maximal cliques whose worst-case runtime is near-optimal, and an experimental evaluation showing the practical utility of the algorithm. Index Terms—Graph Mining, Uncertain Graph, Maximal Clique, Dense Substructure F 1 INTRODUCTION of the most basic problems in graph mining, and has been applied in many settings, including in finding overlapping Large datasets often contain information that is uncertain communities from social networks [14, 17, 18, 19], finding in nature. For example, given people A and B, it may not overlapping multiple protein complexes [20], analysis of be possible to definitively assert a relation of the form “A email networks [21] and other problems in bioinformat- knows B” using available information. Our confidence in ics [22, 23, 24]. such relations are commonly quantified using probability, and we say that the relation exists with a probability of p, for While the notion of a dense substructure and methods some value p determined from the available information.
    [Show full text]
  • Matroid Enumeration for Incidence Geometry
    Discrete Comput Geom (2012) 47:17–43 DOI 10.1007/s00454-011-9388-y Matroid Enumeration for Incidence Geometry Yoshitake Matsumoto · Sonoko Moriyama · Hiroshi Imai · David Bremner Received: 30 August 2009 / Revised: 25 October 2011 / Accepted: 4 November 2011 / Published online: 30 November 2011 © Springer Science+Business Media, LLC 2011 Abstract Matroids are combinatorial abstractions for point configurations and hy- perplane arrangements, which are fundamental objects in discrete geometry. Matroids merely encode incidence information of geometric configurations such as collinear- ity or coplanarity, but they are still enough to describe many problems in discrete geometry, which are called incidence problems. We investigate two kinds of inci- dence problem, the points–lines–planes conjecture and the so-called Sylvester–Gallai type problems derived from the Sylvester–Gallai theorem, by developing a new algo- rithm for the enumeration of non-isomorphic matroids. We confirm the conjectures of Welsh–Seymour on ≤11 points in R3 and that of Motzkin on ≤12 lines in R2, extend- ing previous results. With respect to matroids, this algorithm succeeds to enumerate a complete list of the isomorph-free rank 4 matroids on 10 elements. When geometric configurations corresponding to specific matroids are of interest in some incidence problems, they should be analyzed on oriented matroids. Using an encoding of ori- ented matroid axioms as a boolean satisfiability (SAT) problem, we also enumerate oriented matroids from the matroids of rank 3 on n ≤ 12 elements and rank 4 on n ≤ 9 elements. We further list several new minimal non-orientable matroids. Y. Matsumoto · H. Imai Graduate School of Information Science and Technology, University of Tokyo, Tokyo, Japan Y.
    [Show full text]
  • Implementation Issues of Clique Enumeration Algorithm
    05 Progress in Informatics, No. 9, pp.25–30, (2012) 25 Special issue: Theoretical computer science and discrete mathematics Research Paper Implementation issues of clique enumeration algorithm Takeaki UNO1 1National Institute of Informatics ABSTRACT A clique is a subgraph in which any two vertices are connected. Clique represents a densely connected structure in the graph, thus used to capture the local related elements such as clus- tering, frequent patterns, community mining, and so on. In these applications, enumeration of cliques rather than optimization is frequently used. Recent applications have large scale very sparse graphs, thus efficient implementations for clique enumeration is necessary. In this paper, we describe the algorithm techniques (not coding techniques) for obtaining effi- cient clique enumeration implementations. KEYWORDS maximal clique, enumeration, community mining, cluster mining, polynomial time, reverse search 1 Introduction A maximum clique is hard to find in the sense of time A graph is an object, composed of vertices, and edges complexity, since the problem is NP-hard. Finding a such that each edge connects two vertices. The fol- maximal clique is easy, and can be done in O(Δ2) time lowing is an example of a graph, composed of vertices where Δ is the maximum degree of the graph. Cliques 0, 1, 2, 3, 4, 5 and edges connecting 0 and 1, 1 and 2, 0 can be considered as representatives of dense substruc- and 3, 1 and 3, 1 and 4, 3 and 4. tures of the graphs to be analyzed, the clique enumera- tion has many applications in real-world problems, such as cluster enumeration, community mining, classifica- 0 12 tion, and so on.
    [Show full text]
  • Parametrised Enumeration
    Parametrised Enumeration Arne Meier Februar 2020 color guide (the dark blue is the same with the Gottfried Wilhelm Leibnizuniversity Universität logo) Hannover c: 100 m: 70 Institut für y: 0 k: 0 Theoretische c: 35 m: 0 y: 10 Informatik k: 0 c: 0 m: 0 y: 0 k: 35 Fakultät für Elektrotechnik und Informatik Institut für Theoretische Informatik Fachgebiet TheoretischeWednesday, February 18, 15 Informatik Habilitationsschrift Parametrised Enumeration Arne Meier geboren am 6. Mai 1982 in Hannover Februar 2020 Gutachter Till Tantau Institut für Theoretische Informatik Universität zu Lübeck Gutachter Stefan Woltran Institut für Logic and Computation 192-02 Technische Universität Wien Arne Meier Parametrised Enumeration Habilitationsschrift, Datum der Annahme: 20.11.2019 Gutachter: Till Tantau, Stefan Woltran Gottfried Wilhelm Leibniz Universität Hannover Fachgebiet Theoretische Informatik Institut für Theoretische Informatik Fakultät für Elektrotechnik und Informatik Appelstrasse 4 30167 Hannover Dieses Werk ist lizenziert unter einer Creative Commons “Namensnennung-Nicht kommerziell 3.0 Deutschland” Lizenz. Für Julia, Jonas Heinrich und Leonie Anna. Ihr seid mein größtes Glück auf Erden. Danke für eure Geduld, euer Verständnis und eure Unterstützung. Euer Rückhalt bedeutet mir sehr viel. ♥ If I had an hour to solve a problem, I’d spend 55 minutes thinking about the problem and 5 min- “ utes thinking about solutions. ” — Albert Einstein Abstract In this thesis, we develop a framework of parametrised enumeration complexity. At first, we provide the reader with preliminary notions such as machine models and complexity classes besides proving them to be well-chosen. Then, we study the interplay and the landscape of these classes and present connections to classical enumeration classes.
    [Show full text]
  • Enumeration Algorithms for Conjunctive Queries with Projection
    Enumeration Algorithms for Conjunctive Queries with Projection Shaleen Deep ! Department of Computer Sciences, University of Wisconsin-Madison, Madison, Wisconsin, USA Xiao Hu ! Department of Computer Sciences, Duke University, Durham, North Carolina, USA Paraschos Koutris ! Department of Computer Sciences, University of Wisconsin-Madison, Madison, Wisconsin, USA Abstract We investigate the enumeration of query results for an important subset of CQs with projections, namely star and path queries. The task is to design data structures and algorithms that allow for efficient enumeration with delay guarantees after a preprocessing phase. Our main contribution isa series of results based on the idea of interleaving precomputed output with further join processing to maintain delay guarantees, which maybe of independent interest. In particular, for star queries, we design combinatorial algorithms that provide instance-specific delay guarantees in linear preprocessing time. These algorithms improve upon the currently best known results. Further, we show how existing results can be improved upon by using fast matrix multiplication. We also present new results involving tradeoff between preprocessing time and delay guarantees for enumeration ofpath queries that contain projections. CQs with projection where the join attribute is projected away is equivalent to boolean matrix multiplication. Our results can therefore also be interpreted as sparse, output-sensitive matrix multiplication with delay guarantees. 2012 ACM Subject Classification Theory of computation → Database theory Keywords and phrases Query result enumeration, joins Digital Object Identifier 10.4230/LIPIcs.ICDT.2021.3 Funding This research was supported in part by National Science Foundation grants CRII-1850348 and III-1910014 Acknowledgements We would like to thank the anonymous reviewers for their careful reading and valuable comments that contributed greatly in improving the manuscript.
    [Show full text]
  • Using Binary Decision Diagrams to Enumerate Inductive Logic Programming Solutions
    Using Binary Decision Diagrams to Enumerate Inductive Logic Programming Solutions Hikaru Shindo1, Masaaki Nishino2, and Akihiro Yamamoto1 1 Graduate School of Informatics, Kyoto University, Kyoto, Japan [email protected] 2 NTT Communication Science Laboratories, NTT Corporation, Kyoto, Japan Abstract. This paper proposes a method for efficiently enumerating all solutions of a given ILP problem. Inductive Logic Programming (ILP) is a machine learning technique that assumes that all data, background knowledge, and hypotheses can be represented by first-order logic. The solution of an ILP problem is called a hypothesis. Most ILP studies pro- pose methods that find only one hypothesis per problem. In contrast, the proposed method uses Binary Decision Diagrams (BDDs) to enumerate all hypotheses of a given ILP problem. A BDD yields a very compact graph representation of a Boolean function. Once all hypotheses are enu- merated, the user can select the preferred hypothesis or compare the hy- potheses using an evaluation function. We propose an efficient recursive algorithm for constructing a BDD that represents the set of all hypothe- ses. In addition, we propose an efficient method to get the best hypothesis given an evaluation function. We empirically show the practicality of our approach on ILP problems using real data. Keywords: Inductive Logic Programming · Binary Decision Diagram. 1 Introduction This paper proposes a method that can efficiently enumerate all solutions of a given ILP problem. Inductive Logic Programming (ILP) is a machine learning technique that assuming that each datum, background knowledge, and hypothe- sis can be represented by first-order logic. Our work is based on the foundation of ILP established by Plotkin [17] and Shapiro [21]: inductive inference with clausal logic.
    [Show full text]
  • An Optimal Key Enumeration Algorithm and Its Application to Side-Channel Attacks
    An Optimal Key Enumeration Algorithm and its Application to Side-Channel Attacks Nicolas Veyrat-Charvillon?, Beno^ıtG´erard??, Mathieu Renauld???, Fran¸cois-Xavier Standaerty UCL Crypto Group, Universit´ecatholique de Louvain. Place du Levant 3, B-1348, Louvain-la-Neuve, Belgium. Abstract. Methods for enumerating cryptographic keys based on par- tial information obtained on key bytes are important tools in crypt- analysis. This paper discusses two contributions related to the practical application and algorithmic improvement of such tools. On the one hand, we observe that modern computing platforms al- low performing very large amounts of cryptanalytic operations, approx- imately reaching 250 to 260 block cipher encryptions. As a result, cryp- tographic key sizes for such ciphers typically range between 80 and 128 bits. By contrast, the evaluation of leaking devices is generally based on distinguishers with very limited computational cost, such as Kocher's Differential Power Analysis. We bridge the gap between these cryptana- lytic contexts and show that giving side-channel adversaries some com- puting power has major consequences for the security of leaking devices. For this purpose, we first propose a Bayesian extension of non-profiled side-channel attacks that allows us to rate key candidates in function of their respective probabilities. Next, we investigate the impact of key enumeration taking advantage of this Bayesian formulation, and quantify the resulting reduction in the data complexity of the attacks. On the other hand, we observe that statistical cryptanalyses usually try to reduce the number and size of lists corresponding to partial informa- tion on key bytes, in order to limit the time and memory complexity of the key enumeration.
    [Show full text]
  • Measuring the Complexity of Join Enumeration in Query Optimization Kiyoshi Ono I, Guy M
    Measuring the Complexity of Join Enumeration in Query Optimization Kiyoshi Ono I, Guy M. Lehman IRM Almaden Research Center KS5/80 I, 650 Harry Road San ,Josc, CA 95120 A b&act Introduction Since relational database management systems typically A query optimizer in a relational DRMS translates support only diadic join operators as primitive opera- non-procedural queries into a pr0cedura.l plan for ex- tions, a query optimizer must choose the “best” sc- ecution, typically hy generating many alternative plans, quence of two-way joins to achieve the N-way join of estimating the execution cost of each, and choosing tables requested by a query. The computational com- the plan having the lowest estimated cost. Increasing plexity of this optimization process is dominated by this set offeasilile plans that it evaluates improves the the number of such possible sequences that must bc chances - but dots not guarantee! - that it will find evaluated by the optimizer. This paper describes and a bcttct plan, while increasing the (compile-time) cost measures the performance of the Starburst join enu- for it to optimize the query. A major challenge in the merator, which can parameterically adjust for each design of a query optimizer is to ensure that the set of query the space of join sequences that arc evaluated feasible plans contains cflicient plans without making by the optimizer to allow or disallow (I) composite the :set too big to he gcncratcd practically. tables (i.e., tables that are themselves the result of a join) as the inner operand of a join and (2) joins between two tables having no join predicate linking them (i.e., Cartesian products).
    [Show full text]
  • On Solving Optimization Problems Using Boolean Satisfiability
    Proceeding of the First International Conference on Modeling, Simulation and Applied Optimization, Sharjah, U.A.E. February 1-3, 2005 ON SOLVING OPTIMIZATION PROBLEMS USING BOOLEAN SATISFIABILITY Fadi A. Aloul American University of Sharjah Department of Computer Engineering P.O. Box 26666, Sharjah, UAE [email protected] ABSTRACT SAT solvers require that the problem be represented in CNF form. The last few years have seen significant advances in Boolean sat- While this is applicable to some Engineering tasks, it poses a sig- isfiability (SAT) solving. This has lead to the successful deploy- nificant obstacle to many others. In particular to tasks that need to ment of SAT solvers in a wide range of problems in Engineering express “counting constraints” which impose a lower or upper and Computer Science. In general, most SAT solvers are applied to bound on a certain number of objects. Expressing such constraints Boolean decision problems that are expressed in conjunctive nor- in CNF cannot be efficiently done. Recently, SAT solvers were ex- mal form (CNF). While this input format is applicable to some en- tended to handle pseudo-Boolean (PB) constraints which can easi- gineering tasks, it poses a significant obstacle to others. One of the ly represent “counting constraints” [5, 6, 10, 16, 27]. PB main advances in SAT is generalizing SAT solvers to handle stron- constraints are more expressive and can replace an exponential ger representation of constraints. Specifically, pseudo-Boolean number of CNF constraints [5]. Besides expressing Boolean deci- (PB) constraints which are efficient in representing counting con- sion problems, a key advantage of using PB constraints is the abil- straints and can replace an exponential number of CNF constraints.
    [Show full text]
  • Enumerating Geometric and Combinatorial Objects by Reverse Search
    Enumerating geometric and combinatorial objects by reverse search Shin-ichi Tanigawa RIMS, Kyoto University November 10, 2011 1 Reverse search 1.1 How to design an enumeration algorithm Let V be a finite set V with n = jV j. We shall consider the problem of enumerating a family O of subsets of V which have a specified combinatorial structure. For example, V = f1; 2; : : : ; ng and O is the set of all permutations, or V is an edge set of a graph and O is a family of (edge sets of) spanning trees in the graph. Below we list the procedure for enumerating O by reverse search. Step 1: Define a connected search graph on O. Step 1.1: Define a flip operation, • that is, a (simple) operation which generates a new object o1 2 O from the current object o2 2 O. E.g., swap operations on permutations; diagonal flips on triangulations. • A flip operation is called a k-flip if the symmetric difference between o1 and o2 is at most 2k. Step 1.2: Define the search graph G = (O;E) • where vertices are corresponding to the objects and o1 and o2 are adjacent iff they can be converted to each other by a single flip. Step 1.3: Prove G is connected. Step 2: Define a search tree in G. Step 2.1: Define a unique root of G and the parent for each non-root node of G. • The parent should be defined by using only the local information. • The function which returns a parent is called a parent function.
    [Show full text]
  • Generalizing Boolean Satisfiability II
    Journal of Artificial Intelligence Research 22 (2004) 481-534 Submitted 08/04; published 12/04 Generalizing Boolean Satisfiability II: Theory Heidi E. Dixon [email protected] On Time Systems, Inc. 1850 Millrace, Suite 1 Eugene, OR 97403 USA Matthew L. Ginsberg [email protected] On Time Systems, Inc. 1850 Millrace, Suite 1 Eugene, OR 97403 USA Eugene M. Luks [email protected] Computer and Information Science University of Oregon Eugene, OR 97403 USA Andrew J. Parkes [email protected] CIRL 1269 University of Oregon Eugene, OR 97403 USA Abstract This is the second of three planned papers describing zap, a satisfiability engine that substantially generalizes existing tools while retaining the performance characteristics of modern high performance solvers. The fundamental idea underlying zap is that many problems passed to such engines contain rich internal structure that is obscured by the Boolean representation used; our goal is to define a representation in which this structure is apparent and can easily be exploited to improve computational performance. This paper presents the theoretical basis for the ideas underlying zap, arguing that existing ideas in this area exploit a single, recurring structure in that multiple database axioms can be obtained by operating on a single axiom using a subgroup of the group of permutations on the literals in the problem. We argue that the group structure precisely captures the general structure at which earlier approaches hinted, and give numerous examples of its use. We go on to extend the Davis-Putnam-Logemann-Loveland inference procedure to this broader setting, and show that earlier computational improvements are either subsumed or left intact by the new method.
    [Show full text]